A new baseline for edge detection: Make encoder–decoder great again

Yachuan Li

doi:10.1016/j.image.2026.117485

A new baseline for edge detection: Make encoder–decoder great again

2026-01-14

0

PRE

AI

DOI：10.1016/j.image.2026.117485

Original

Original request for help

Save

Abstract

En 中文

• Huge computational cost and complex strategy hinder the progress of edge detection. • A vanilla and pure encoder–decoder based edge detection baseline is needed. • NBED decouples the extraction process of spatial features and semantic features. • Cascaded high-resolution feature decoder can suppress the noise and the spatial error. • Powerful encoder–decoders can disable the feature augmentation and fusion modules.

A new baseline for edge detection: Make encoder–decoder great again

Abstract

Journal

Organization